🧠 Customer Sentiment Watchdog — Hackathon Project 🧠

🚀 Problem Statement:

Customer Sentiment Watchdog for Support Teams**

"It’s hard for CX teams to detect rising frustration or satisfaction trends across support channels (emails, chats, tickets)."

💡 Our Solution:

We built an AI-powered dashboard that analyzes customer feedback in real-time (from web forms, emails, and tickets), classifies emotional tone (positive, neutral, negative), and alerts the team if negative sentiment spikes.

Our virtual AI assistant, LIZA, provides real-time suggestions based on recent complaints using LLMs (via Sarvam AI/OpenRouter API).

---

🛠️ Key Features

✅ Sentiment analysis from feedback
✅ Visual analytics (charts for sentiment, activity, top complaints)
✅ Chat-based assistant “LIZA” for suggestions
✅ Live feedback stream
✅ Tracks engagement levels: Promoters, Passives, Detractors
✅ Clean Tailwind + Chart.js frontend
✅ Full Flask backend with CSV-based feedback storage

---

🤖 Tech Stack

• Frontend: HTML, TailwindCSS, Chart.js
• Backend: Python, Flask
• AI Integration: Sarvam AI via OpenRouter API (used in suggestion_agent.py)
• Data Handling: pandas, CSV
• Hosting Ready: Portable for deployment on Render/Heroku/Localhost

---

💬 What We Learned

This was a 3-day intense hackathon where:

• We used AI responsibly — yes, we took help, but we engineered prompts, debugged countless errors, iterated fast, and learned deeply.
• We explored prompt-tuning, model handling, backend integration, and frontend polish.
• I made mistakes, learned from them, and gave this project my 💯 percent.
• The biggest takeaway: how to use AI tools efficiently to solve real problems.

We hope you like our project as much as we loved building it. 💜

---

📦 Folder Structure

📁 project-root │ ├── app.py # Main Flask server ├── feedback.csv # Stores all feedback data ├── suggestion_agent.py # AI generation using Sarvam AI ├── sentiment_agent.py # Handles sentiment classification ├── feedback_handler.py # Handles data saving ├── chart_generator.py # Prepares data for frontend │ ├── templates/ │ ├── insights.html # Dashboard (main) │ ├── index.html # Web feedback form │ ├── ticket.html # Ticket feedback form │ └── email.html # Email feedback form

yaml Copy Edit

---

👥 Team Contribution

This project is my solo submission for a hackathon, built with:

AI support (Sarvam, ChatGPT, etc.)

My own full-stack implementation

3 days of hands-on effort, learning, debugging, and improving every module

🌟 Final Words

This was not just a project, but a learning journey. From prompt engineering to building a full-stack AI tool, this experience helped me grow immensely as a developer and a problem-solver.

Thank you for checking it out. ❤️

📋 Requirements

Install these using pip install -r requirements.txt

Flask
pandas
requests

# Step 1: Clone this repo
git clone https://github.com/MeNoodie/customer_sentiment.project.git
cd sentiment-watchdog

# Step 2: Create virtual env (optional)
python -m venv venv
source venv/bin/activate  # or venv\Scripts\activate on Windows

# Step 3: Install dependencies
pip install -r requirements.txt

# Step 4: Run the app
python app.py


headers = {
    "Authorization": "Bearer YOUR_OPENROUTER_API_KEY",
    "Content-Type": "application/json"
}



Let me know if you want a badge section (`Built With`, `Hackathon Submission`, etc.) or a zipped release for submission.